Fetching the paper…

Frame Context Packing and Drift Prevention in Next-Frame-Prediction Video Diffusion Models · Around